New version of the AI for Good chapter

[marcozennaro]

No description provided.

[profvjreddi]

Thanks for the contribution! 👍

[happyappledog]

it's very hard to find on-device medical AI cases. I am amazed you found these examples!
I found another on-device example if you want to broaden the use cases (they are not necessary for low resource tho): Implantable cardioverter-defibrillator with on-device ML to detect ventricular arrhythmia

https://youtu.be/vx2gWzAr85A?t=2359

[happyappledog]

https://www.nature.com/articles/s42256-023-00659-9

Ventricular fibrillation and ventricular tachycardia are life-threatening ventricular arrhythmias (VAs) and the primary causes of sudden cardiac death, resulting in significant morbidity and mortality1. Individuals at high risk of sudden cardiac death rely on implantable cardioverter–defibrillators (ICDs) to provide timely and appropriate defibrillation treatment in case of life-threatening VAs1. However, existing industry practice is simple rule-based detection methods, which have not been updated over the past few decades2,

TDC’22 designated the NUCLEO-L432KC development kit (STMicroelectronics) as the targeting MCU platform for all participating teams. This $10 development board is equipped with an ARM Cortex-M4 core at 80 MHz, 256 kB of flash memory and 64 kB of SRAM, and its power consumption is around 30 mW in operation and 1.5 mW when idling. The IEGMs data were collected and provided by Singular Medical using ICDs... With the training dataset and unified evaluation platform, participating teams could utilize either the existing frameworks or their own tools to develop, train and deploy the AI/ML algorithm on board with cross-layer optimizations.

[profvjreddi]

Emma, cool. Thanks! Do you think you could help add this idea of competitions into the material and link to this work please?

…

On Sun, Nov 5, 2023 at 10:40 PM happyappledog ***@***.***> wrote: https://www.nature.com/articles/s42256-023-00659-9 TDC’22 designated the NUCLEO-L432KC development kit (STMicroelectronics) as the targeting MCU platform for all participating teams. This *$10 development* board is equipped with an ARM Cortex-M4 core at 80 MHz, 256 kB of flash memory and 64 kB of SRAM, and its power consumption is around 30 mW in operation and 1.5 mW when idling. The IEGMs data were collected and provided by Singular Medical using ICDs... With the training dataset and unified evaluation platform, participating teams could utilize either the existing frameworks or their own tools to develop, train and deploy the AI/ML algorithm on board with cross-layer optimizations. — Reply to this email directly, view it on GitHub <#32 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABT6DFD7TJXMSHFDHYPO3Q3YDBE2JAVCNFSM6AAAAAA6D6LZ2GVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOJUGAZTMNJXG4> . You are receiving this because you modified the open/close state.Message ID: ***@***.***>

-- Vijay Janapa Reddi, Ph. D. | John L. Loeb Associate Professor of Engineering and Applied Sciences | John A. Paulson School of Engineering and Applied Sciences | Science and Engineering Complex (SEC) | 150 Western Ave, Room #5.305 | Boston, MA 02134 | Harvard University | Email ***@***.***> | Website <http://scholar.harvard.edu/vijay-janapa-reddi> | Google Scholar <https://scholar.google.com/citations?hl=en&user=gy4UVGcAAAAJ&view_op=list_works&sortby=pubdate> | Edge Computing Lab <https://edge.seas.harvard.edu> | Schedule a Meeting <https://scholar.harvard.edu/vijay-janapa-reddi/schedule> | Admin <https://scholar.harvard.edu/vijay-janapa-reddi/contact> |