Across multiple disciplines, artificial intelligence (AI) in medical screening devices is helping improve patient outcomes by offering healthcare practitioners more nuanced ways to interpret data and make clinical decisions. AI assistance can play a critical role in detecting cardiovascular diseases earlier, which leads to earlier treatment and more precise outcome prediction and prognosis evaluation.
AI accomplishes this through a technique called machine learning by quickly analyzing massive amounts of data and sharing clinically relevant insights. This use of statistical algorithms identifies patterns in new data that match the patterns that have already been learned from existing data. The resulting predictions improves the quality, precision, and efficiency of healthcare.
To understand the different ways machine learning works, these are three subtypes of the technique based on the way the predictive model learns and accumulates data.
Supervised learning includes logistic regression, support-vector machines (SVMs), and neural networks. Models are developed using human-labeled datasets. These models find the most relevant variables to predict or classify future events and outcomes. Each new piece of data trains the predictive model, making it more robust over time.
Unsupervised learning includes cluster analysis. This method identifies new relationships within data by finding hidden structures in datasets that weren’t previously categorized in the training dataset.
Reinforcement learning involves reward-based learning and is typically used in gaming and robotic applications. This predictive model uses tools such as sensors, cameras, and GPS to help the machine understand the surrounding environment. The interactions with the environment guide the machine to continuously learn and improve.
These machine learning methods expand what’s possible in healthcare. For example, supervised learning algorithms can be used to classify heart rhythms. When AI powers an ECG, anomalies are automatically detected through a precise, quantitative assessment of cardiac functions. This removes variabilities that may skew results and reduces the time it takes to interpret the data, making appointments more efficient.
As AI further integrates into the cardiology world, physicians will be more and more empowered to improve their patients’ outcomes. The extra layer of intelligence augments the physician’s work, rather than replacing it. To learn more about how AI-driven ECGs can help detect cardiac abnormalities earlier, see nCardio.
A 12-lead ECG is the gold standard of measuring heart rhythms and spotting irregularities. Six leads are applied to the patient’s chest, while the remaining 6 leads are divided among the patients’ extremities. The heart’s electrical activity generates a current that spreads to the skin. Each lead contains a sensor that measures the electrical activity sent from the sinoatrial node through the heart. The most basic devices simply record this information onto paper. Smart ECGs take this process several steps further by integrating algorithms that help the physician interpret the results.
For example, nCardio’s internally developed filtering and detection algorithms identify and classify various cardiac abnormalities. The reading generates standard and extended measurements that include an extended capture of up to 40 seconds full-disclosure and markings for 15 independent events. Upon completion, human-centered augmented intelligence provides diagnostics and reporting that support clinical decisions.
This combination of a routine ECG with 12 leads along with interpretation and reporting can be billed with CPT code #93000. This code reports both the professional and technical components of the service. The ECG test is typically covered when there are documented signs and symptoms or other clinical indications for providing the service. The average amount that healthcare providers can be reimbursed for the service typically ranges from $10-$20.
Coverage is often only provided when the review and interpretation is completed by a physician and the device is owned by the ordering provider. When submitting for CPT code #93000 reimbursement, remember to include the full graphic tracings and a formal written report with the interpretation and signature from the physician. nCardio provides the tracings and reporting needed for your submissions.
For situations without the interpretation and report, when only the tracing is completed during the electrocardiogram, the CPT code #93005 can be used. The average reimbursement is about $9.
nCardio provides each element that is needed for your insurance reimbursement for both codes. This streamlines the billing process and makes it easier to get reimbursed.
Cervical cancer typically develops slowly over many years, even decades, often without symptoms. The long, asymptomatic development is one of the many reasons why cervical cancer goes undetected and is the fourth most common cancer in women, worldwide. Despite this, cervical cancer is entirely preventable and the eradication of the disease lies in prevention.
Cervical pre-cancer screening is the most powerful form of prevention. The success of these screenings is due, in part, to the repeat testing that women typically undergo over many years.
In Europe and the United States, it’s protocol for screen positive women to be investigated and treated using colposcopy. The conventional colposcope hasn't changed significantly since its introduction nearly 100 years ago in Hamburg. This basic colposcopy allows magnified, illuminated examination of the surface epithelial layers of the ‘at-risk’ tissue area known as the Transformation Zone (TZ). The subjectivity and variability in the results create a larger margin of error.
Smart colposcopes, on the other hand, reduce the subjectivity from conventional devices through augmented intelligence. This form of AI increases precision by amplifying the physician’s skill, rather than replacing it. Objectivity is achieved by adding AI-driven cervical mapping technology that visually quantifies the cellular response to the acetic acid wash. Upon capturing the image, visual inspection algorithms can calculate the speed, intensity, and duration of the acetowhitening effect within seconds.
The result is cervical mapping, which looks similar to a heatmap. This visualization draws the physician’s attention to any potentially abnormal cells that may require a biopsy. Because cervical mapping instantly displays exam results, women can be screened and if necessary, treated in a single visit. The practitioner can immediately interpret and deliver the results, determine next steps, and use the enhanced biopsy selection if needed.
This streamlined, digital screening makes it possible for cervical cancer screening to be more accessible and precise for all women globally. Through regular screenings, cervical pre-cancer can be caught and treated in the earliest stages before it develops into cancer. This is how we can eradicate the disease worldwide.
To learn more about how cervical mapping works and how it transforms cervical pre-cancer screening, download the whitepaper.
Typically a part of the biannual teeth cleaning and exam, oral cancer screening is essential for identifying disease at the earliest stages, before the patient has any symptoms. Conventionally, this procedure is performed without any screening devices or augmented intelligence, which can result in overlooked disease or false positives. Using a tool with augmented intelligence, however, can not only deliver precise and objective results, but it also can be reimbursed by insurance companies.
nOra is an oral cancer screening device that uses intraoral imaging sensors that push beyond conventional white light to more precisely detect pre-cancerous and early-stage lesions. The USB-based peripheral with specialized illumination includes white and narrow band sources, customized ISP, and low light sensors. With it, tiny lesions appear black, showing a greater sensitivity and intensity than conventional white lights. Internally developed AI and NLP algorithms identify and classify cancerous & pre-cancerous lesions, providing immediate results.
While the dental hygienist may perform the screening with the device, the dentist must also evaluate and provide a final determination of the presented condition and diagnosis to report the service for reimbursement. Insurance companies typically require dentists to substantiate the oral cancer screening claim by submitting clinical notes that detail the visit, along with the resulting images. At the end of the procedure, nOra provides a full pdf report that can be sent to the insurance company as proof of the completed service.
Using a device for the oral cancer screening exam is a valuable and billable procedure that can cost the patient $25 to $65. The fee charged is a practice decision that is typically based on reimbursement.
The American Dental Association (ADA) assigned the CDT-5 code D0431 for insurance reimbursement. It is for a pre-diagnostic test using a screening device and not the actual biopsy procedure, which is a different code. It should only be billed in addition to the dentist’s completed physical and visual exams on the patient.
The CDT-5 code D0431 is described as “adjunctive pre-diagnostic oral cancer screening that aids in detection of mucosal abnormalities including premalignant and malignant lesions, not to include cytology or biopsy procedures.” Note that using a device like nGyn is a separate procedure beyond the conventional device-free oral cancer screening that’s typically included in the oral exam. This is why a screening service performed with a device requires an additional code.
In addition to the ADA’s code D0431, a dental practice might also be able to seek reimbursement from medical insurance. Dentists can bill medical insurance for some medically necessary dental procedures that screen for medical conditions. To meet the coding requirements to bill medical insurance for oral cancer screenings, a screening device must be used; it cannot be a visual evaluation only.
Two codes need to be used when billing medical insurance for an oral cancer screening. The first is the CPT code, or the procedure code, which is 82397 for chemiluminescent assay. Then the ICD-10 code or diagnosis code is Z12.81 for screening for malignant neoplasm of the oral cavity. These two codes, along with the procedure report, tell the medical insurance the patient’s medical story. While not all medical insurance companies will reimburse for these codes, many do.
Oral cancer screening is a potentially life saving service, allowing early treatment for the disease before the patient has symptoms. With the increased precision and objectivity that smart cancer screening devices bring, it is beneficial to offer the service to adult patients to accurately catch disease in the earliest pre-cancer stages.