Healthcare Software: Treatment Planning
Using Software Tools to Plan Patient Care
Developing treatment and procedural plans plays a crucial part in efficient, accurate, and quality healthcare. Researchers at Emory University have worked to develop several planning models, software, and algorithms to personalize and improve many medical procedures – and elevate patient experience.
Imaging Innovations
Accurate Prostate Cancer Imaging for High-Dose-Rate Brachytherapy Planning
High-dose-rate (HDR) brachytherapy, which delivers radiotherapeutics to the prostate through small, inserted catheters, is a common treatment for patients suffering from prostate cancer. Accurate catheter placement is essential for optimal treatment, but current methods to identify potential, effective catheter sites are not always reliable. Emory researchers are addressing this problem with a treatment planning model that uses 3-D intra-operative ultrasound (TRUS) images, which can show the prostate after catheter insertions and post-operative CT scans. This new planning method allows doctors to use the more accurate 3D volume of a prostate to increase treatment accuracy and decrease the risk of affecting neighboring healthy tissue.
High-Resolution Field Inhomogeneity Correction for Magnetic Resonance Spectroscopy
Magnetic resonance spectroscopy (MRS) is a non-invasive technique to detect metabolic changes in the tissues of the human body. Chemical exchange saturation transfer (CEST) is an MR technique that enables imaging of certain compounds, like glutamate and phosphocreatine, at concentrations that are often too low to be imaged directly. Although CEST MRI has entered clinical use, its contrast is limited, and intravoxel inhomogeneity continues to pose a challenge that current algorithms cannot resolve in either routine MRS or CEST spectroscopy. However, Emory researcher Phillip Zhe Sun, MD, has developed a high-resolution field map-based deconvolution algorithm for the correction of intravoxel field inhomogeneity in MR spectroscopy. The CEST intravoxel inhomogeneity correction (CIVIC) approach may assist in imaging certain compounds for the diagnosis of diseases.
Imaging Software for Individualized Treatment with Radiotherapy
Prostate cancer is the most common cancer in American men, and current radiotherapy methods risk missing treatment margins or delivering unnecessary radiation because the prostate can shift position between sessions. A shift of mere millimeters can make a big difference. However, a new predictive mathematical model developed at Emory offers patient‑specific radiation margins and correctly predicted prostate motion for 91.7% of the 22 patients. This technology has the potential to significantly improve the accuracy and safety of image‑guided radiotherapy.
AI Tools for Better Diagnoses
Machine Learning Diagnostic for Automated Identification and Classification of Bone Marrow Cells
Each year, more than half a million U.S. patients undergo bone marrow aspiration biopsies, in which liquid marrow is extracted and smeared onto a glass slide for microscopic examination. This process can help diagnose hematologic diseases such as anemia, leukemia, lymphoma, and multiple myeloma, but the process is slow and labor‑intensive, highlighting the need for faster, more accurate automated assessment methods. Emory inventors have developed a more precise and efficient bone marrow analysis tool that uses artificial intelligence trained on whole-slide digital pathology images of marrow aspirate smears to accurately identify 16 key bone marrow cell types.
Personalized Brachytherapy Planning: Knowledge-Based Algorithm Adapts to Physician Preferences
Existing planning software for low-dose brachytherapy (LDR) – one of the most common treatments for early-stage prostate cancer – doesn't use a knowledge-based treatment planning platform. But Emory researchers have developed a treatment planning software tailored to physician preferences for prostate brachytherapy, starting with previous physician-specific therapy plans. With those preferences, the algorithm compares data from past cases to the current case in order to form an LDR plan that is optimal for both the patient and the physician. This software can help standardize plan quality of LDR treatment delivery while simultaneously decreasing planning time.
Smarter Charts
Software for Reviewing Patients in Radiotherapy Chart Rounds
Communication about a patient’s status is usually done through a manual weekly chart review of the patient's treatment record. But this method is often laborious, inefficient, and easy to delay the detection of a potential treatment error. However, Emory inventors have developed a software system to automate chart checks, streamline peer-review for radiotherapy patients, and use artificial intelligence to help check treatment accuracy. The software offers a quality assurance platform for chart rounds review and has been used at Emory University since 2016.