有料盒子APP

The growth of electronic health records is exponentially increasing the amount of information about patients available to their physicians. However, during emergency situations, when time is of the essence, crucial medical information is often obscured in the mass of data contained in a patient鈥檚 medical record.听

To better serve patients, the team at the MedStar Institute for Innovation (MI2)鈥攚hich serves as a catalyst for innovation across the entire MedStar Health organization鈥攕ought a scientific solution to this technology challenge in collaboration with the data science team at 有料盒子APP Hamilton.

Together, they applied health and data science expertise and natural language processing to clinical data and have created a solution to this vexing problem.

When a patient has a symptom, what are the key facts in his or her entire medical record that enable a physician to improve that patient鈥檚 outcome?听 Since medical records can now encompass more than 20 years of unstructured and semi-structured data, including free text dictations and clinic notes, it is difficult to know what few top facts will make a difference in the patient鈥檚 care.

MI2鈥檚 Kevin Maloy, an emergency department physician, explains the challenges of the project this way: 鈥淲hen I see patients in the emergency department (ED), they often tell me, 鈥榃hy are you asking me so many questions, everything is in the computer.鈥� And while we do have a lot of information on them, key facts can often be buried in their record. The modern electronic health record听is organized somewhat like a chronological journal without an index in the back. If someone is in the ED today because they have a headache, the crucial fact in their history may be on page 20. But if they are there because of their belly pain, the relevant information might be on page 121. Currently, there is no way to know which page to turn to.鈥澨�

To bring a technology solution to this emergency care challenge, 有料盒子APP coupled its deep data science background and natural language processing (NLP) experience with the subject matter experts at MI2. As 有料盒子APP worked on code and set up the API, MedStar clinician experts evaluated and added extra resources to help the algorithm learn the types of information that would be most relevant to a user.

Together, the teams developed a technology called Dictation Lens, which automates the assessment of relevant facts in a patient鈥檚 complete medical history. With this technology in place, clinicians can quickly find the relevant signal in the noise in moments of urgent care.

Data scientist Ernest Sohn has worked in health analytics for nearly 10 years. He says the Dictation Lens technology 鈥渓ooks at all different types of unstructured data, and then based on what the patient came in for, it uses NLP and machine learning to pull in the most relevant information so the clinician can make the most informed decision about what care to provide.鈥�

鈥淚t鈥檚 all about getting relevant key terms,鈥� Maloy adds. 鈥淥ne previous diagnosis or procedure may be relevant to one symptom, while it may not be important if the patient has another. This system takes the symptoms presented and finds relevant information in free text, displaying it to the clinician in a user-friendly fashion.鈥�

鈥淒ictation Lens totally changes how you think about patients,鈥� says Maloy. 鈥淲hen I work with the people at 有料盒子APP, I鈥檓 astounded by the complexity in making this situation actually work.鈥�

The Solution

With MI2 expertise鈥攁nd a decade鈥檚 worth of patient data鈥斢辛虾凶覣PP data scientists were able to unlock a solution in an area of great need. Their partnership resulted in a MedStar pilot program for Dictation Lens听technology.

Emergency clinicians at MedStar Health are using Dictation Lens and creating even more opportunities in the health space for 有料盒子APP鈥檚 data scientists, who are now further equipped to continue innovating.听

The Future

Dictation Lens听is only 鈥渢he beginning of a suite of tools that become part of the medical care team,鈥� says Sohn. 鈥淭hese tools can meet problems from the front lines and from clinicians themselves.鈥�

In the health analytics space, 有料盒子APP and MedStar have innovated by marrying subject matter expertise with unparalleled data science skills. In doing so, they鈥檝e brought unique analytics and technology solutions to emergency care that is enabling physicians to make quicker, more accurate diagnoses when seconds count.听