From our Expert Voices conversation on AI and medicine.

Making the right diagnosis is a huge part of getting quality of care right, and it has not gotten enough attention. Diagnosis is one of the most complex aspects of the medical interaction: beginning with honest, comprehensive and detailed information from the patient, astute observation and physical examination, leading to intelligently selected imaging and/or laboratory tests and finally interpreted through knowledge, analysis, judgement and intuition.

Even the smartest, most well-trained physician has patterns of thinking that can lead her to jump to conclusions. Doctors have human brains like everyone else, and AI can help as a super smart, super fast consultant at your side. Medical schools should teach doctors (and nurses) to partner intelligently with information technology. It will soon be as essential as anatomy and physiology.

Yes, but: Don't forget the patient. The X-rays and labs don't always give a clear, unambiguous answer. Diagnosis is often a process, not an event. And patients know their symptoms, they know what's changing, they can give clues that the doctor or the computer might not think of. The doctor and the computer both need to listen to the patient.

The bottom line: Just like you sometimes know more about local circumstances than Google maps, the patient brings essential personal and situational information to the analysis. The clinician, the patient and the computer need to collaborate to get to the right answer in the most effective way.

Other voices in the conversation:

• Jack Stockert, Managing Director, Health2047: A doctor's helper, but not a doctor
• Eric Topol, professor of molecular medicine, Scripps Research Institute: Improving the doctor-patient relationship
• Ethan Weiss, associate professor, UC-San Francisco School of Medicine: Helping doctors make better diagnoses