I have blogged previously about the use of algorithms in healthcare which will be revolutionary in terms of diagnosing patients and even predicting which diseases they may develop in the future (see, for example: Eric Schmidt Discusses the Potential Value of Predictive Analytics in the ER; An Algorithm Using Medical Record Data Predicts Risk for Parkinson's Disease). A recent article discussed how radical this change will be (see: How Health Care Changes When Algorithms Start Making Diagnoses). Needless to say, some politicians are already making foolish judgements about medical algorithms as quoted in the following excerpt:
...[A] team at Google used data on eye scans from over 125,000 patients to build an algorithm that could detect retinopathy, the number one cause of blindness in some parts of the world, with over 90% accuracy, on par with board-certified ophthalmologists...[T]hese results had the same constraints [as similar other AI studies]; humans could not always fully comprehend why the models made the decisions they made....Earlier this year, France’s minister of state for the digital sector flatly stated that any algorithm that cannot be explained should not be used. But opposing these advances wholesale is not the answer. The benefits of an algorithmic approach to medicine are simply too great to ignore. Earlier detection of ailments like skin cancer or cardiovascular disease could lead to reductions in morbidity thanks to these methods. Poorer economies with limited access to trained physicians may benefit as well, as a host of diseases may be found and treated earlier. Individualized treatment recommendations may also improve, leading to saved lives for some and increased quality of life for many others.
Here's a quote from the Wikipedia about deep learning and neural networks to frame this blog note (see: Deep learning):
Deep learning...is part of a broader family of machine learning methods based on learning data representations, as opposed to task-specific algorithms. Learning can be supervised, semi-supervised or unsupervised. Deep learning architectures such as deep neural networks, deep belief networks and recurrent neural networks have been applied to fields including computer vision, speech recognition, natural language processing, audio recognition, social network filtering, machine translation, bioinformatics and drug design, where they have produced results comparable to and in some cases superior to human experts.
I was struck by the inanity of the observation by France’s minister of state that "any algorithm that cannot be explained should not be used." There have been numerous observations over past years that have saved lives but couldn't be initially explained in scientific terms. Penicillin, a miracle cure in its day, is only one such example (see: Discovery and Development of Penicillin). Algorithms are already in use that can predict the future development of a disease without necessarily understanding all of the factors that came into play in terms of making such a prediction. Predictive algorithms are always tested in the "real world" concerning their validity. Deep learning and neural networks are just another type of scientific tool that will advance medicine and should not be singled out as somehow mysterious and therefore suspect.