What do you think?
Rate this book
The Hidden Life of the Basal Ganglia: At the Base of Brain and Mind
The anatomy and physiology of the basal ganglia and their relation to brain and behavior, disorders and therapies, and philosophy of mind and moral values.
The main task of the basal ganglia—a group of subcortical nuclei, located at the base of the brain—is to optimize and execute our automatic behavior. In this book, Hagai Bergman analyzes the anatomy and physiology of the basal ganglia, discussing their relation to brain and behavior, to disorders and therapies, and even to moral values. Drawing on his forty years of studying the basal ganglia, Bergman presents new information on physiology and computational models, Parkinson’s disease and other ganglia-related disorders, and such therapies as deep brain stimulation.
Focusing on studies of nonhuman primates and human basal ganglia and relying on system physiology and in vivo extra-cellular recording techniques, Bergman first describes the major brain structures that constitute the basal ganglia, the morphology of their cellular elements, their synaptic connectivity and their physiological function in health and disease. He discusses the computational physiology of the healthy basal ganglia, describing four generations of computational models, and then traces the computational physiology of basal ganglia–related disorders and their treatments, including Parkinson’s disease and its pharmacological and surgical therapies. Finally, Bergman considers the implications of these findings for such moral concerns as free will. Explaining this leap into domains rarely explored in neuroscientific accounts, Bergman writes that the longer he studies the basal ganglia, the more he is convinced that they are truly the base of both brain and mind.
The main task of the basal ganglia—a group of subcortical nuclei, located at the base of the brain—is to optimize and execute our automatic behavior. In this book, Hagai Bergman analyzes the anatomy and physiology of the basal ganglia, discussing their relation to brain and behavior, to disorders and therapies, and even to moral values. Drawing on his forty years of studying the basal ganglia, Bergman presents new information on physiology and computational models, Parkinson’s disease and other ganglia-related disorders, and such therapies as deep brain stimulation.
Focusing on studies of nonhuman primates and human basal ganglia and relying on system physiology and in vivo extra-cellular recording techniques, Bergman first describes the major brain structures that constitute the basal ganglia, the morphology of their cellular elements, their synaptic connectivity and their physiological function in health and disease. He discusses the computational physiology of the healthy basal ganglia, describing four generations of computational models, and then traces the computational physiology of basal ganglia–related disorders and their treatments, including Parkinson’s disease and its pharmacological and surgical therapies. Finally, Bergman considers the implications of these findings for such moral concerns as free will. Explaining this leap into domains rarely explored in neuroscientific accounts, Bergman writes that the longer he studies the basal ganglia, the more he is convinced that they are truly the base of both brain and mind.
332 pages, Paperback
Published October 12, 2021
7 people are currently reading
114 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 3 of 3 reviews
October 4, 2025
This book offers a unique perspective on the basal ganglia and parkinson's disease from an electrophysiology point of view. I've read countless articles regarding neuromodulation for parkinson's disease and a textbook on microelectrode recording for deep brain stimulation. However, this book provides a unique set of information not commonly found elsewhere. The devil is in the details, and this book uncovers the details regarding all the major nuclei in the basal ganglia.
The book is highly technical and abstract. The author has a distinctive tone in his writing, often striving to be pithy. Unfortunately, he fails. It only reads as an unnecessary distraction. Furthermore, the author attempts to compact rigorous technical facts into mere sentences or small paragraphs. At times, the author attempts to get philosophical with the neuroscience, which I found moot. I wish it read like a textbook because then it would be more structured and straightforward. The diagrams and scientific findings had poor quality and really small text, unlike a textbook. I found two-thirds of the book to be highly pragmatic and practical. The rest of the book is highly theoretical or philosophical.
Part I: Introduction & Background
These chapters were rich. Although the diagrams were poor in quality, I got a lot out of this section. The author did a great job in tying together all the structures in the basal ganglia, comparing and contrasting each structure (caudate, putamen, substantia niagra, etc.). Additionally, the author spends a lot of time piecing together the nuclei of the thalamus and their synaptic connections to the basal ganglia, cerebellum, and motor cortex.
The cell physiology was well written for every nucleus in the basal ganglia. The author discusses how these cells connect to other structures, their electrophysiological signature, and their role in modulating movement. Additionally, we understand the neurotransmitters involved with these structures. It was interesting to learn that the subthalamic nucleus is the only nucleus in the basal ganglia that is glutamatergic. The rest of the nuclei are inhibitory (GABA). Lastly, every major structure in the basal ganglia is a simple one-layer network, composed of many projection neurons. Hence, there are no local, complex microcircuits as seen in the cortex and cerebellum.
Chapters 5 & 6 describe how to quantify slow and fast electrical behavior in neurons; local field potentials, and single unit activity, respectively. Additionally, we are provided with an overview of an introperative neurophysiological monitoring system. The author briefly mentions the basics of grounding, amplifiers, filtering, and the frequency domain.
Part II: Computational Physiology of the Healthy Basal Ganglia
Chapter 7 discusses the discharge patterns for all the major structures in the basal ganglia and how that relates to their physiology. The most scintillating nucleus to learn about was the Gpe and its "pausers" displayed from the recording system. The most interesting fact regarding this chapter is that the Gpe is the main output structure for the basal ganglia.
Chapters 9 & 10 discuss the box and arrow models (D1/D2) for the basal ganglia. I was already familiar with these models, but the author added more context. Additionally, the author added an anatomical update by incorporating the PPN (pedunculopontine nucleus) into the model. Fascinating. I had just recently learned about the heavy impact this nucleus plays, along with the substania niagra, for coordinating and executing gait.
Part III: Computational Physiology of Basal Ganglia Disorders & Their Therapy
Chapter 13 provides a shining light on movement disorders. More specifically, the author compares and contrasts the differences between extrapyramidal and pyramidal disorders. Extrapyramidal diseases come from the cerebellum and basal ganglia. Abnormal movements, changes in muscle tone, postural abnormalities, and tremor characterize these disorders. Furthermore, we can have hypokinetic or hyperkinetic disorders, although there is a fuzzy line between the two. As we progress further into the chapter, a broad overview of the pathophysiology of parkinson's is described.
Chapter 14 provides a brief overview of the impact that animal models have had on understanding Parkinson's disease. I really appreciated the author's ethical views and responsibilities regarding experimentation on live animals.
Chapters 15 & 16 provide further details regarding spiking patterns, LFPs, and beta oscillations.
Given my clinical background in DBS, chapters 18 & 19 were highly relevant to me.
Ch 18 was a highly comprehensive chapter regarding DBS. The author sprinkles fruitful aspects of the DBS procedure, microelectrode recording, programming, stereotaxy, but most importantly, the hypothesized mechanisms for DBS.
The author stipulates that DBS acts as an "information lesion". Lesioning for DBS procedures (ablation) was more often performed in the past, but they were effective. The lesion is clearly having an impact on a node, similar to an open circuit. Hence, the pathological information is blocked. With DBS, injecting current into the area has a clear effect on the symptoms, but we don't know why. Popular theory among neurology, neurosurgery, and medical device representatives states that the stimulation is activating the basal ganglia. Due to the lack of dopamine, there is less activation in these areas. However, the author discusses synaptic depletion. Brilliant. Furthermore, you can only stimulate a neuron so many times before its intracellular vesicles, which house the neurotransmitters, are completely depleted. Once there are no more vesicles, it has no means to communicate with the neighboring synapse. DBS is causing synaptic failure, which has a functional inactivation effect similar to ablation therapy. However, according to the author, the stimulation rate has to be above 80 hertz for the information lesion to take effect. In conclusion, "stimulation" is misleading, at least as it relates to the mechanism of DBS.
Ch19 centered around closed-loop DBS. The author advocates for closed-loop DBS while at the same time claiming, "Neurophysiology is still an art, and most importantly, it does not pretend to be perfect”. It makes no sense advocating for closed loop while the electrophysiology is highly suspect and too theoretical for clinical application. According to the author, as it relates to the spectrum of LFPs that act as biomarkers for Parkinson's disease, "Even the phase-specific approach is naive because it divides the brain waves into two classes, good and bad, and assumes that all Parkinson's disease symptoms are the same". I would have to agree. Lastly, the beta band is only clinically useful for rigidity. This assumes you are even able to obtain a high-quality recording. I will end this note by stating that low beta is "likely associated" with akinetic rigid symptoms and that high beta is "potentially associated" with akinetic rigid symptoms. These are statements taken away from Medtronic's IFU for their brain-sensing platform.
The book is highly technical and abstract. The author has a distinctive tone in his writing, often striving to be pithy. Unfortunately, he fails. It only reads as an unnecessary distraction. Furthermore, the author attempts to compact rigorous technical facts into mere sentences or small paragraphs. At times, the author attempts to get philosophical with the neuroscience, which I found moot. I wish it read like a textbook because then it would be more structured and straightforward. The diagrams and scientific findings had poor quality and really small text, unlike a textbook. I found two-thirds of the book to be highly pragmatic and practical. The rest of the book is highly theoretical or philosophical.
Part I: Introduction & Background
These chapters were rich. Although the diagrams were poor in quality, I got a lot out of this section. The author did a great job in tying together all the structures in the basal ganglia, comparing and contrasting each structure (caudate, putamen, substantia niagra, etc.). Additionally, the author spends a lot of time piecing together the nuclei of the thalamus and their synaptic connections to the basal ganglia, cerebellum, and motor cortex.
The cell physiology was well written for every nucleus in the basal ganglia. The author discusses how these cells connect to other structures, their electrophysiological signature, and their role in modulating movement. Additionally, we understand the neurotransmitters involved with these structures. It was interesting to learn that the subthalamic nucleus is the only nucleus in the basal ganglia that is glutamatergic. The rest of the nuclei are inhibitory (GABA). Lastly, every major structure in the basal ganglia is a simple one-layer network, composed of many projection neurons. Hence, there are no local, complex microcircuits as seen in the cortex and cerebellum.
Chapters 5 & 6 describe how to quantify slow and fast electrical behavior in neurons; local field potentials, and single unit activity, respectively. Additionally, we are provided with an overview of an introperative neurophysiological monitoring system. The author briefly mentions the basics of grounding, amplifiers, filtering, and the frequency domain.
Part II: Computational Physiology of the Healthy Basal Ganglia
Chapter 7 discusses the discharge patterns for all the major structures in the basal ganglia and how that relates to their physiology. The most scintillating nucleus to learn about was the Gpe and its "pausers" displayed from the recording system. The most interesting fact regarding this chapter is that the Gpe is the main output structure for the basal ganglia.
Chapters 9 & 10 discuss the box and arrow models (D1/D2) for the basal ganglia. I was already familiar with these models, but the author added more context. Additionally, the author added an anatomical update by incorporating the PPN (pedunculopontine nucleus) into the model. Fascinating. I had just recently learned about the heavy impact this nucleus plays, along with the substania niagra, for coordinating and executing gait.
Part III: Computational Physiology of Basal Ganglia Disorders & Their Therapy
Chapter 13 provides a shining light on movement disorders. More specifically, the author compares and contrasts the differences between extrapyramidal and pyramidal disorders. Extrapyramidal diseases come from the cerebellum and basal ganglia. Abnormal movements, changes in muscle tone, postural abnormalities, and tremor characterize these disorders. Furthermore, we can have hypokinetic or hyperkinetic disorders, although there is a fuzzy line between the two. As we progress further into the chapter, a broad overview of the pathophysiology of parkinson's is described.
Chapter 14 provides a brief overview of the impact that animal models have had on understanding Parkinson's disease. I really appreciated the author's ethical views and responsibilities regarding experimentation on live animals.
Chapters 15 & 16 provide further details regarding spiking patterns, LFPs, and beta oscillations.
Given my clinical background in DBS, chapters 18 & 19 were highly relevant to me.
Ch 18 was a highly comprehensive chapter regarding DBS. The author sprinkles fruitful aspects of the DBS procedure, microelectrode recording, programming, stereotaxy, but most importantly, the hypothesized mechanisms for DBS.
The author stipulates that DBS acts as an "information lesion". Lesioning for DBS procedures (ablation) was more often performed in the past, but they were effective. The lesion is clearly having an impact on a node, similar to an open circuit. Hence, the pathological information is blocked. With DBS, injecting current into the area has a clear effect on the symptoms, but we don't know why. Popular theory among neurology, neurosurgery, and medical device representatives states that the stimulation is activating the basal ganglia. Due to the lack of dopamine, there is less activation in these areas. However, the author discusses synaptic depletion. Brilliant. Furthermore, you can only stimulate a neuron so many times before its intracellular vesicles, which house the neurotransmitters, are completely depleted. Once there are no more vesicles, it has no means to communicate with the neighboring synapse. DBS is causing synaptic failure, which has a functional inactivation effect similar to ablation therapy. However, according to the author, the stimulation rate has to be above 80 hertz for the information lesion to take effect. In conclusion, "stimulation" is misleading, at least as it relates to the mechanism of DBS.
Ch19 centered around closed-loop DBS. The author advocates for closed-loop DBS while at the same time claiming, "Neurophysiology is still an art, and most importantly, it does not pretend to be perfect”. It makes no sense advocating for closed loop while the electrophysiology is highly suspect and too theoretical for clinical application. According to the author, as it relates to the spectrum of LFPs that act as biomarkers for Parkinson's disease, "Even the phase-specific approach is naive because it divides the brain waves into two classes, good and bad, and assumes that all Parkinson's disease symptoms are the same". I would have to agree. Lastly, the beta band is only clinically useful for rigidity. This assumes you are even able to obtain a high-quality recording. I will end this note by stating that low beta is "likely associated" with akinetic rigid symptoms and that high beta is "potentially associated" with akinetic rigid symptoms. These are statements taken away from Medtronic's IFU for their brain-sensing platform.
December 12, 2022
I enjoyed the author's witty remarks and occasional connections to the big questions and philosophy. I also liked the qualitative approach the author undertook in estimating the number of neurons and their projections. It was surprising to see how much experimental work validates the reinforcement-learning model of the basal ganglia. Overall, it is a great book on the summary of what the basal nuclei do.
Less exciting (at least for me) were the chapters about Parkinson's disease and experimental methods since I am a purely computational person. Sometimes the author also goes talking about issues too distant from the topic at hand (for instance, his story about how he discovered the existence of the dimensionality reduction idea).
Here is a quick summary of the core facts about the BG:
The BG came in with the appearance of the vertebrates.
Anatomically, the BG comprises of Striatum, Globus Pallidus, Substantia Nigra, and Subthalamic Nucleus.
Striatum:
-Dorsal Striatum: caudate nucleus and putamen
-Ventral Striatum: Nucleus Accumbens (NAc)
Globus pallidus (pale clump!)
-External (GPe)
-Internal (GPi)
-Ventral pallidum
Substantia nigra (black substance!)
-Pars reticulata (SNr)
-Pars compacta (SNc)
-Ventral tegmental area (VTA)
Subthalamic nucleus (STN)
All cortical areas bilaterally topographically project to the striatum - the major input nucleus of the BG. STN receives input from the ipsilateral motor and somatosensory cortices.
GPi and SNr (major output nuclei) project to the thalamus and to the frontal cortex.
About half of the glutamatergic innervation of the striatum comes from the thalamus, while another half comes from the cortex. Among the core structure of the BG, only the STN is glutamatergic, the other structures are GABAergic.
The striatum mainly contains GABAergic medium spiny neurons (MSNs, 80-95%) which are innervated by dopaminergic axons from SNc (to dorsal striatum) and VTA (to ventral striatum). The striatum, like all the other structures in BG, is a one-layer nucleus.
There are two populations of striatal MSNs: D1 and D2 MSNs, named after the predominant dopaminergic receptor on the soma.
D1 MSNs project to the output structures of the BG (GPi and SNr) - direct pathway.
D2 MSNs project to GPe - indirect pathway.
5-20% of the striatal neurons are interneurons, being medium aspiny neurons. Some of the striatal interneurons are cholinergic (1% of the total population of the striatum).
The subthalamic nucleus (STN) receives glutamatergic input from the frontal cortex and thalamus and GABAergic inputs from GPe. It sends glutamatergic projections to the GPe, the GPi and the SNr.
GPe neurons innervate all basal ganglia structures: the striatum, STN, GPi and SNr. Most of the synaptic input (~90%) to GPe, GPi, SNr is coming from GABAergic striatum neurons.
On the first order of approximation, the BG network may be described as a feedforward monolayer network (as opposed to the multilayered and reciprocal cortex connectivity).
Striatal projections are converging: there are fewer neurons in each subsequent layer.
The STN projections, on the other hand, are divergent. In the striatum, MSNs discharge at around 1-2 Hz, and tonically active cholinergic neurons (TANs) discharge at 3-7 Hz.
The discharge rate of the neurons in STN is about 20-30 Hz.
GPe, GPi and SNr continuously discharge with 40-50 Hz (GPi, SNr > GPe).
The discharge of GPe neurons is tonic but with abrupt cessation (~0.6 sec, 10 per minute) and resuscitation of the firing. The distribution of these intervals is Poisson-like, and there is a strong relationship between the interval statistics and an arousal level (decreasing with deep sleep). Why GPe neurons pause is still a mystery. GPe neurons could be classified as high-frequency (>20 Hz) and low-frequency (<20 Hz).
The activity of the neurons in the GPe, GPi, and the SNr are independent of one another and uncorrelated.
The spontaneous spiking in the striatal and SNc dopaminergic neurons is, on the other hand, highly correlated.
The classical model for motor vigor control:
Cortex excites striatum. Dopaminergic neurons of SNc excite D1 and inhibit D2 MSNs.
Excitation of D1 MSNs leads to inhibition of GPi/SNr (direct GABAergic pathway).
Inhibition of D2 MSNs leads to excitation of the GPe, which inhibits the STN. Inhibition of the STN leads to reduced excitation of the GPi/SNr (indirect pathway)
Overall, the increase in dopamine in the striatum leads (directly and indirectly) to GPi/SNr inhibition.
Inhibition of GPi/SNr (output structures of the BG) leads to disinhibition of the thalamocortical motor areas, which control movements.
Another function of the BG is a specific action selection.
The selection of the specific movements requires activation of the direct pathway, while the inhibition of competing movements requires activation of the indirect pathway.
A reinforcement-learning take on the BG is that it implements actor-critic architecture and that the dopamine signal encodes temporal difference error. A burst of dopaminergic neurons encodes states in which reality is better than expected. Thus, the SNc constitutes a critic: the neurons in SNc homogeneously encode prediction error, they widely broadcast to the striatum. The dopaminergic neurons implement so-called “volume transmission”: each neuron gives rise to ~0.5 million terminals, and not every one of these terminals is opposed by postsynaptic boutons - volume transmission.
Dopamine modules the efficacy of corticostriatal projections. Form the book, I have gotten an impression that so far it is the most successful model, accounting for multiple experimental facts, however, there are still open questions:
How to encode the negative rewards with only positive firing rates?
How to account for multiobjective optimization?
The multiobjective optimization could be somehow connected to the fact that there are multiple neurotransmitters at play in the SNc: dopamine, serotonin, acetylcholine, and histamine, which orchestrate the striatal activity.
The SNc may be controlling both the motor vigor in general and the specific action selection by 1) controlling the excitability of MSNs in the striatum 2) changing the efficacy of the corticospinal synapses.
The author also suggests that the BG and cerebellum may be the neural basis of the unconscious Kahneman’s system 1, while the frontal cortex is a rational system 2.
Less exciting (at least for me) were the chapters about Parkinson's disease and experimental methods since I am a purely computational person. Sometimes the author also goes talking about issues too distant from the topic at hand (for instance, his story about how he discovered the existence of the dimensionality reduction idea).
Here is a quick summary of the core facts about the BG:
The BG came in with the appearance of the vertebrates.
Anatomically, the BG comprises of Striatum, Globus Pallidus, Substantia Nigra, and Subthalamic Nucleus.
Striatum:
-Dorsal Striatum: caudate nucleus and putamen
-Ventral Striatum: Nucleus Accumbens (NAc)
Globus pallidus (pale clump!)
-External (GPe)
-Internal (GPi)
-Ventral pallidum
Substantia nigra (black substance!)
-Pars reticulata (SNr)
-Pars compacta (SNc)
-Ventral tegmental area (VTA)
Subthalamic nucleus (STN)
All cortical areas bilaterally topographically project to the striatum - the major input nucleus of the BG. STN receives input from the ipsilateral motor and somatosensory cortices.
GPi and SNr (major output nuclei) project to the thalamus and to the frontal cortex.
About half of the glutamatergic innervation of the striatum comes from the thalamus, while another half comes from the cortex. Among the core structure of the BG, only the STN is glutamatergic, the other structures are GABAergic.
The striatum mainly contains GABAergic medium spiny neurons (MSNs, 80-95%) which are innervated by dopaminergic axons from SNc (to dorsal striatum) and VTA (to ventral striatum). The striatum, like all the other structures in BG, is a one-layer nucleus.
There are two populations of striatal MSNs: D1 and D2 MSNs, named after the predominant dopaminergic receptor on the soma.
D1 MSNs project to the output structures of the BG (GPi and SNr) - direct pathway.
D2 MSNs project to GPe - indirect pathway.
5-20% of the striatal neurons are interneurons, being medium aspiny neurons. Some of the striatal interneurons are cholinergic (1% of the total population of the striatum).
The subthalamic nucleus (STN) receives glutamatergic input from the frontal cortex and thalamus and GABAergic inputs from GPe. It sends glutamatergic projections to the GPe, the GPi and the SNr.
GPe neurons innervate all basal ganglia structures: the striatum, STN, GPi and SNr. Most of the synaptic input (~90%) to GPe, GPi, SNr is coming from GABAergic striatum neurons.
On the first order of approximation, the BG network may be described as a feedforward monolayer network (as opposed to the multilayered and reciprocal cortex connectivity).
Striatal projections are converging: there are fewer neurons in each subsequent layer.
The STN projections, on the other hand, are divergent. In the striatum, MSNs discharge at around 1-2 Hz, and tonically active cholinergic neurons (TANs) discharge at 3-7 Hz.
The discharge rate of the neurons in STN is about 20-30 Hz.
GPe, GPi and SNr continuously discharge with 40-50 Hz (GPi, SNr > GPe).
The discharge of GPe neurons is tonic but with abrupt cessation (~0.6 sec, 10 per minute) and resuscitation of the firing. The distribution of these intervals is Poisson-like, and there is a strong relationship between the interval statistics and an arousal level (decreasing with deep sleep). Why GPe neurons pause is still a mystery. GPe neurons could be classified as high-frequency (>20 Hz) and low-frequency (<20 Hz).
The activity of the neurons in the GPe, GPi, and the SNr are independent of one another and uncorrelated.
The spontaneous spiking in the striatal and SNc dopaminergic neurons is, on the other hand, highly correlated.
The classical model for motor vigor control:
Cortex excites striatum. Dopaminergic neurons of SNc excite D1 and inhibit D2 MSNs.
Excitation of D1 MSNs leads to inhibition of GPi/SNr (direct GABAergic pathway).
Inhibition of D2 MSNs leads to excitation of the GPe, which inhibits the STN. Inhibition of the STN leads to reduced excitation of the GPi/SNr (indirect pathway)
Overall, the increase in dopamine in the striatum leads (directly and indirectly) to GPi/SNr inhibition.
Inhibition of GPi/SNr (output structures of the BG) leads to disinhibition of the thalamocortical motor areas, which control movements.
Another function of the BG is a specific action selection.
The selection of the specific movements requires activation of the direct pathway, while the inhibition of competing movements requires activation of the indirect pathway.
A reinforcement-learning take on the BG is that it implements actor-critic architecture and that the dopamine signal encodes temporal difference error. A burst of dopaminergic neurons encodes states in which reality is better than expected. Thus, the SNc constitutes a critic: the neurons in SNc homogeneously encode prediction error, they widely broadcast to the striatum. The dopaminergic neurons implement so-called “volume transmission”: each neuron gives rise to ~0.5 million terminals, and not every one of these terminals is opposed by postsynaptic boutons - volume transmission.
Dopamine modules the efficacy of corticostriatal projections. Form the book, I have gotten an impression that so far it is the most successful model, accounting for multiple experimental facts, however, there are still open questions:
How to encode the negative rewards with only positive firing rates?
How to account for multiobjective optimization?
The multiobjective optimization could be somehow connected to the fact that there are multiple neurotransmitters at play in the SNc: dopamine, serotonin, acetylcholine, and histamine, which orchestrate the striatal activity.
The SNc may be controlling both the motor vigor in general and the specific action selection by 1) controlling the excitability of MSNs in the striatum 2) changing the efficacy of the corticospinal synapses.
The author also suggests that the BG and cerebellum may be the neural basis of the unconscious Kahneman’s system 1, while the frontal cortex is a rational system 2.
April 28, 2025
When I first got this book It looked really daunting, I had already the previous book in the series by Gareth Leng but this book was a bit bigger so thought it would be a bit of a slog
Bergman's style of writing is very engaging and it's clear his ideas come from diverse sources, concepts in this book are well explained and even revisited which was appreciated. Enjoyed the bullet point summaries at the end of each chapter to see what Bergman wanted you to take away really helped me separate key info from not-so-key info. However this book has a heavy emphasis on computational neuro, I enjoyed these chapters but my background in this field is lacking so struggled with these chapters. felt like the BG was quite mysterious prior to reading so enjoyed discovering what its functions are how it is organised and what disorders/diseases are centred around the BG.
Bergman's style of writing is very engaging and it's clear his ideas come from diverse sources, concepts in this book are well explained and even revisited which was appreciated. Enjoyed the bullet point summaries at the end of each chapter to see what Bergman wanted you to take away really helped me separate key info from not-so-key info. However this book has a heavy emphasis on computational neuro, I enjoyed these chapters but my background in this field is lacking so struggled with these chapters. felt like the BG was quite mysterious prior to reading so enjoyed discovering what its functions are how it is organised and what disorders/diseases are centred around the BG.
Displaying 1 - 3 of 3 reviews