The New Doing Gooder Google

January 17, 2020

Google’s cheerleading unit likes to remind us, amid the constant criticisms, that the company makes some positive contributions to society. For example, it seems their AI has gotten good at detecting cancer. We learn from AndoridCentral that “Google’s AI Is Better at Detecting Cancer than Doctors, Says Study.” About the same research, Ausdroid reports, “Google Publish their Impressive Breast Cancer Screening Using AI Results.” The capabilities are courtesy of technology developed by Google acquisition DeepMind. The study was performed by Google Health in conjunction with Cancer Research UK Imperial Centre, Northwestern University, and Royal Surrey County Hospital. Researchers used deep-learning tools to create AI detection models and applied them to almost 30,000 patients for whom results were already known. Muhammad Jarir Kanji of AndroidCentral writes:

“The system was trained using a large dataset of mammograms from women in the two countries. Even more telling than its better accuracy than doctors was the fact that it did so with far less information than the radiologists it was competing with, who also had access to the patients’ medical history and previous mammograms in their deliberations. … While the paper noted that ‘AI may be uniquely poised to help with’ the challenge of detecting breast cancer, Darzi said the system was not yet at a stage where it could replace a human reader.”

Emphasis on “yet.” Meanwhile, Ausdroid’s Scott Plowman emphasizes:

“The data sets were also NOT used to train the AI system and thus we totally unknown to the system.

Comparing the positive results from the AI to those patients who ended up having biopsy-confirmed breast cancer the AI demonstrated a ‘statistically significant’ improvement in ‘absolute specificity’ of 1.2% (UK – double read), and 5.7% (USA – single read) and an improvement in absolute sensitivity of 2.7% (UK) and 9.4% (USA). For reference, sensitivity is the ability to correctly identify lesions and specificity is how accurate it is at identifying those without lesions. This means that it has a reduction in both false positives and false negatives.”

If Google’s PR team spins more stories like this one, they just might be able to burnish the company’s reputation.

Cynthia Murrell, January 08, 2020

Comments

Got something to say?





  • Archives

  • Recent Posts

  • Meta