Hairuo industry large model rides the cloud in the medical industry

 “Hello, Mr. Li, what is wrong with you?”

“I’ve been feeling a lot of stomachaches lately.”

“Where does it hurt? Does it hurt here? ”

“Yes, it hurts when you press it.”

”Did you take any medicine?”

……

Thousands of patients visit the hospital every year, and in every hospital, large or small, such conversations are frequent.

During patients’ hospitalization, in order to fully grasp information such as patients’ physical state and medical history, the doctor needs to communicate with patients frequently, and form medical records such as admission records and ward round records. As the core of medical information, the accuracy and completeness of medical records are crucial to patient treatment and hospital management, so filing this part of the medical record paperwork will consume a lot of time and energy of medical staff, which not only greatly increases the workload of doctors, but also may lead to the omission or error of certain information due to various uncertain factors.

In the First Affiliated Hospital of Shandong First Medical University (Shandong Provincial Qianfoshan Hospital), located in Jinan, it is easier for doctors to do ward rounds because of the emergence of the Hairuo medical large model. By identifying doctor-patient dialogues, the Hairuo medical large model can complete the automatic generation of medical records in a few seconds, which greatly improves the work efficiency of doctors. “In practical applications, the accuracy rate of generating electronic medical records through Hairuo can reach about 90%.” A doctor in a department said.

Based on the in-depth study of massive medical data, the Hairuo medical large model can realize the automatic generation of electronic medical records through five links: speech recognition, analysis and transcription, medical record generation, medical record quality control, and confirmation and import, so as to ensure the accuracy and completeness of medical record generation. Through dual-role speech recognition technology, the Hairuo medical large model can automatically separate the voice of doctors and patients, and support the transformation of voice information such as Mandarin and dialects into professional medical terms. Through semantic analysis technology, the Hairuo medical large model can also intelligently extract key consultation information, standardize and structure it, output the diagnosis process in real time, and automatically fill it into the medical template to automatically generate electronic medical records. At the same time, the Hairuo medical large model can analyze the format, completeness and accuracy of the medical record through the embedded electronic medical record quality control system, and medical records can be imported into the hospital’s medical record system after a simple review and modification by the doctor.

This process can reduce the workload of doctors by 70%.

In addition, the Hairuo industry large model can also be combined with professional private data to enhance training, and be customized according to the specific needs of hospitals and doctors to meet the needs of medical record generation in different scenarios.

In addition to the generation of electronic medical records, Inspur Cloud is also actively promoting the implementation of Hairuo Medical Large Model Agents such as assisted triage and intelligent pre-consultation in Qianfoshan Hospital.

Relying on in-depth learning and natural language processing technology, the Hairuo medical large model has the ability of intelligent sub-guidance, and through natural dialogue with patients, the Hairuo medical large model can accurately and quickly identify the patient’s condition, and automatically recommend the appropriate departments and doctors, with a recommendation accuracy rate of more than 95%. At the same time, the Hairuo medical large model can also provide intelligent Q&A and decision support services, which can provide preliminary medical advice and guidance according to the patient’s symptoms and information.

In addition, the Hairuo medical large model also innovatively applies the interactive experience mode of multi-modal and online and offline interaction, making the medical treatment process more intuitive and convenient. Before registering online, patients can use mobile phones and other devices to communicate with Hairuo through voice, text and other ways to obtain triage recommendations; When seeking medical treatment offline, patients can have real-time conversations with the interactive large screen with built-in Hairuo virtual human medical guides, and obtain services such as in-hospital navigation. This greatly improves the patient’s medical experience and relieves the work pressure of the guides.

In terms of intelligent pre-consultation, the Hairuo medical large model adopts multi-modal large model and natural language processing technology, supports text, voice, picture input and other forms, can imitate the doctor’s consultation logic, and carefully inquire about patients from multiple dimensions such as disease course, symptom severity, potential triggers, medication history and allergy history, and dynamically adjust follow-up questions according to patient feedback to obtain the most complete disease information. In the end, the Hairuo medical large model will send the patient’s medical record to the doctor’s computer, and the patient’s medical record information can be directly seen by clicking on it, so that the doctor’s time on providing medical services can be “well spent”.

Behind the efficiency improvement is the complete product support capability of Hairuo. Supported by a distributed computing platform, based on the basic capabilities such as data management, model development, and cluster management of the Hairuo industry large model, and relying on massive medical data, the Hairuo medical large model is gradually growing into a medical large model that promotes the intelligent transformation of the medical industry.

For patients, the emergence of the Hairuo medical large model can largely solve the pain points encountered by patients in the process of medical treatment, and bring patients a more efficient and high-quality medical experience.

For hospitals, empowered by the Hairuo industry large model, doctors work more efficiently, and hospitals also discover potential disease trends and better diagnosis and treatment models through the mining and analysis of a large number of medical record data, thus optimizing resource allocation and improving the quality of medical services.

The story of Qianfoshan Hospital is also being played out in other hospitals and medical institutions...