Tech & Innovation in Healthcare

Medical image analysis can help detect other conditions, too.

With the pandemic wreaking havoc across the US, healthcare providers started deploying various health IT solutions to address emerging challenges. They aimed to reduce the load on doctors and protect clinicians and patients from the virus exposure, which allowed providers to continue to deliver due care and avoid closure.

According to the American Healthcare Association (AHA) report, shutdowns became a painful reality for about 8 percent of the respondents, which is equal to 16,000 private practices nationwide (URL: www.aha.org/aha-center-health-innovation-market-scan/2020-09-01-specialist-and-private-practices-take-severe). Moreover, over 75 percent of them were specialists.

Use AI for Accurate Medical Image Analysis

During the acute phase of the pandemic, healthcare image analysis became a key assistant in diagnosing coronavirus-related pneumonia. Well-trained deep learning models helped practitioners save time, performing analysis sometimes 10 times faster than human experts. Moreover, AI-driven tools of this kind also show higher diagnostic accuracy.

For example, the medical image analysis solution developed at the University of Wuhan outperformed professional radiologists. Chinese specialists created an ML-driven system analyzing critical patient data, such as age, gender, and more, and CT scans. When the researchers compared the system’s performance with that of radiologists, they found the tool detected COVID-19 infection in about 70 percent of patients whose CT scans medical experts considered normal. As a result, deep learning models can find complex patterns the human eye can’t perceive.

Deep learning models can also analyze chest X-rays successfully. Though the accuracy of such tools is up to 86 percent, with adequate training, they can reach the accuracy of the CT scans analysis. Given that an X-ray scanner offers higher portability and lower cost, the current level of accuracy is acceptable. Such ML-based solutions may help clinicians diagnose the disease earlier, administer treatment on time, and avoid comorbidities.

Other applications: Providers can apply medical image analysis to several other specialties as well. AI-driven analysis can assist doctors with pulmonary function testing (PFT) for chronic obstructive pulmonary disease (COPD) diagnostics and risk prognostication, diabetic retinopathy and tuberculosis detection, and diagnosing diverse cancer types, including melanoma, breast, and prostate cancer. Additionally, adequately trained deep learning models can read ultrasound, microscopy, endoscopy, and other types of medical images.

Treat Patients Safely with Telemedicine

Deploying telemedicine solutions allowed providers to continue offering remote consultations. Patients picked up the idea enthusiastically — in one year, the share of telemedicine consultations tripled, from 20 percent in March 2020 to 60 percent in March 2021 (URL: www.statista.com/statistics/1256460/use-of-telehealth-telemedicine-appointments-in-the-us/).

To help doctors with telemedicine tech adoption, specialists from the Stanford University School of Medicine prepared a guide. The guide concisely describes five steps a typical telehealth visit contains, from preparations doctors need to take before each consultation to monitoring a patient’s home environment and their emotional state. Besides, the federal government took the necessary measures to motivate doctors to use telemedicine tools by introducing regulations ruling that teleconsultations and regular in-house visits receive equal reimbursement.

Staying power: As the situation with COVID-19 improved, the number of telemedicine visits has reduced. However, the new care provision model gained a strong foothold in providers’ services. Telehealth visits have become a good alternative to some in-house therapies.

Telemedicine has proved efficient for managing some behavioral issues, such as anxiety and depression. The technology also allows providers to administer physiotherapy for diverse musculoskeletal conditions, pain management, and injury rehabilitation. Additionally, telemedicine became a standard offer for some nursing homes, sparing immobilized residents the need to go to offline consultations.

Employ Remote Patient Monitoring for Complex Patients

During the pandemic, caring for patients with chronic conditions increased in complexity. Fortunately, providers have had an ally at their disposal — remote patient monitoring (RPM) technologies.

Using specific devices, patients collect their health data and send the information securely to their providers. In the event a parameter drops or rises to a critical level, the tool notifies the doctor, the patient’s relatives, and the emergency care unit. This care model allowed patients and providers to minimize human contact, thus lowering potential virus exposure.

Mixed reception: While select chronic-condition patients view the tools favorably, the devices aren’t meeting the expectations of others. The reasons vary — some patients have difficulty managing the device, others can’t enjoy the solutions due to the cost or connectivity issues.

To deliver care to those left behind, providers started offering device-free patient monitoring. In this case, doctor-patient communication occurs over a mobile or landline phone connection. Every week, medical professionals prepare questionnaires to evaluate a patient’s health status. When a patient fills in the questionnaire, their doctor calls them to explain the results and consult on further actions.

Today, behavioral health makes the most popular choice for device-free patient monitoring. However, Esse Health, a provider-patient community, has leveraged this model for patients suffering from heart failure, COPD, and diabetes, too.

Revolutionize Patient Care After the Pandemic

Automated image analysis, telemedicine, and RPM have helped providers deliver care effectively amidst the pandemic. These solutions allowed them to reduce workload and keep some patients safely at home, thus lowering the risk of exposure for both parties.

Providers continue to adapt the new care models to patients’ needs. Image analysis and telemedicine are aiding an increasing number of fields, while patients in remote and rural locations experience improved access to care thanks to patient monitoring. Plus, given the providers’ flexible approach and attention to patients, these technologies have established a foothold in the industry and will evolve over time.