BRN (ASX) Unlimited Use Cases

Originally posted by Fact Finder: ↑

Brand new:

License: CC BY 4.0
arXiv:2507.15958v1 [eess.IV] 21 Jul 2025
Quantization-Aware Neuromorphic Architecture for Efficient Skin Disease Classification on Resource-Constrained Devices
Haitian Wang12, Xinyu Wang2, Yiren Wang2, Karen Lee2, Zichen Geng2,
Xian Zhang2, Kehkashan Kiran1, Yu Zhang1†, Bo Miao3†Corresponding author: Yu Zhang. Contact: [email protected]Mailing address (China): 1 Dongxiang Road, Chang’an District, Xi’an, Shaanxi 710129, P.R. China. Phone: (+86) 138919975111Northwestern Polytechnical University, Xi’an, Shaanxi 710129, China2The University of Western Australia, Perth, WA 6009, Australia3Australian Institute for Machine Learning, University of Adelaide, SA 5005, Australia
Abstract
Accurate and efficient skin lesion classification on edge devices is critical for accessible dermatological care but remains challenging due to computational, energy, and privacy constraints. We introduce QANA, a novel quantization-aware neuromorphic architecture for incremental skin lesion classification on resource-limited hardware. QANA effectively integrates ghost modules, efficient channel attention, and squeeze-and-excitation blocks for robust feature representation with low-latency and energy-efficient inference. Its quantization-aware head and spike-compatible transformations enable seamless conversion to spiking neural networks (SNNs) and deployment on neuromorphic platforms. Evaluation on the large-scale HAM10000 benchmark and a real-world clinical dataset shows that QANA achieves 91.6% Top-1 accuracy and 82.4% macro F1 on HAM10000, and 90.8% Top-1 accuracy and 81.7% macro F1 on the clinical dataset, consistently outperforming leading CNN-to-SNN models under fair comparison. Deployed on BrainChip Akida hardware, QANA achieves 1.5 ms inference latency and 1.7 mJ energy per image, reducing inference latency and energy use by over 94.6%/98.6% compared to GPU-based CNNs, and exceeding the performance of advanced CNN-to-SNN conversion methods. These results demonstrate the effectiveness of QANA for accurate, real-time, and privacy-sensitive medical analysis in edge environments

VConclusion
In this paper, we proposed QANA, a quantization-aware neuromorphic framework for skin lesion classification on edge devices. Extensive experiments on the large-scale HAM10000 benchmark and a real-world clinical dataset show that QANA achieves state-of-the-art accuracy (91.6% Top-1, 82.4% macro F1 on HAM10000; 90.8%/81.7% on the clinical set) while enabling real-time and energy-efficient inference on the BrainChip Akida platform (1.5 ms latency, 1.7 mJ per image). These results demonstrate that QANA is highly effective for portable medical analysis and AI deployment in dermatology under limited computing resources.
VIAcknowledgement
This research was supported by the National Natural Science Foundation of China (Nos. 62172336 and 62032018). The authors gratefully acknowledge BrainChip Holdings Ltd. for providing technical support and the Akida AKD1000 hardware platform, whose powerful neuromorphic computing capabilities enabled strong performance of the SNN model. The authors also extend their appreciation to Dr. Atif Mansoor, Dr. Bo Miao and their teams for their preliminary contributions to this research
The use cases for Brainchip’s AKIDA Technologies is only limited by the imaginations of engineers, data scientists and entrepreneurs.

My opinion only DYOR

Fact Finder

Expand

https://medium.com/@bradleysusser/your-phone-could-soon-diagnose-skin-cancer-thanks-to-a-new-kind-of-ai-f63533cecd38

No.	Vol.	Price($)
75	3157327	19.5¢

No.

Vol.

Price($)

3157327

19.5¢

Price($)	Vol.	No.
20.0¢	323257	12

Price($)

Vol.

No.

20.0¢

323257

No.	Vol.	Price($)
75	3145462	0.195
106	2592837	0.190
37	1788021	0.185
58	1104875	0.180
22	490014	0.175

No.

Vol.

Price($)

3145462

0.195

106

2592837

0.190

1788021

0.185

1104875

0.180

490014

0.175

Price($)	Vol.	No.
0.200	308034	14
0.205	1257424	18
0.210	1490125	23
0.215	912821	18
0.220	1230196	28

Price($)

Vol.

No.

0.200

308034

0.205

1257424

0.210

1490125

0.215

912821

0.220

1230196

(20min delay)
Last 19.8¢		Change -0.008(3.66%)		Mkt cap ! $405.0M

Open	High	Low	Value	Volume
20.0¢	20.5¢	19.5¢	$992.4K	4.947M

Unlimited Use Cases, page-2

Why a Fed rate cut could obliterate your super – and no one's warning you

'Rising demand': Lakes Blue Energy starts well-timed Wombat-5 gas drilling

Prospect Resources: Replicating the success of Arcadia with ~1b tonne copper ore target

Infini Resources appoints Rohan Bone as CEO ahead of Phase 2 campaign

Week 31 Wrap: Oz core CPI @ 2.7%! But here come tariffs. Again. Next Friday

Featured News

Buyers (Bids)

Sellers (Offers)

Featured News

Unlimited Use Cases, page-2

Top Stories

Why a Fed rate cut could obliterate your super – and no one's warning you

'Rising demand': Lakes Blue Energy starts well-timed Wombat-5 gas drilling

Prospect Resources: Replicating the success of Arcadia with ~1b tonne copper ore target

Infini Resources appoints Rohan Bone as CEO ahead of Phase 2 campaign

Week 31 Wrap: Oz core CPI @ 2.7%! But here come tariffs. Again. Next Friday

Featured News

Buyers (Bids)

Sellers (Offers)

Featured News