Otitis media (OM) is a common disease associated with high antibiotics prescription, high recurrence, and developmental issues in children. Its early detection is crucial to prevent sequelae. However, the diagnosis of OM is commonly based on subjective conclusions taken from observation of an eardrum through a conventional otoscope. Therefore, this often leads to misdiagnosis and moreover culminates in erroneous antibiotics prescription, occasioning the appearance of resistant bacteria. A smartphone-based imaging system allows an untrained user to acquire and transmit data to a specialist for remote diagnosis. We thus developed a smartphone-based multispectral imaging otoscope capable of offering quantitative information on the physiological state of an eardrum with the benefits from the portability and connectivity of the smartphone. The system consists of an Android application for the control of the hardware and also the display of classification results, a circuit board for the interface of an LED multiplexer with the smartphone, and a custom-made otoscope probe for uniform illumination onto the interior of an ear canal. The probe includes a set of lenses and eight optical fibers attached to the LED multiplexer. The multiplexer is composed of a white LED and eight LEDs with consecutive/sequential wavelengths. We examined a normal ear and an ear with OM with effusion using our developed system. The results showed that the smartphone-based multispectral imaging otoscope could quantitatively distinguish between a healthy ear and an ear with OM with effusion, suggesting its potential as a mobile healthcare tool for diagnosis and management of middle ear pathologies.
To date, the incident rates of various skin diseases have increased due to hereditary and environmental factors including stress, irregular diet, pollution, etc. Among these skin diseases, seborrheic dermatitis and psoriasis are a chronic/relapsing dermatitis involving infection and temporary alopecia. However, they typically exhibit similar symptoms, thus resulting in difficulty in discrimination between them. To prevent their associated complications and appropriate treatments for them, it is crucial to discriminate between seborrheic dermatitis and psoriasis with high specificity and sensitivity and further continuously/quantitatively to monitor the skin lesions during their treatment at other locations besides a hospital. Thus, we here demonstrate a mobile multispectral imaging system connected to a smartphone for selfdiagnosis of seborrheic dermatitis and further discrimination between seborrheic dermatitis and psoriasis on the scalp, which is the more challenging case. Using the system developed, multispectral imaging and analysis of seborrheic dermatitis and psoriasis on the scalp was carried out. It was here found that the spectral signatures of seborrheic dermatitis and psoriasis were discernable and thus seborrheic dermatitis on the scalp could be distinguished from psoriasis by using the system. In particular, the smartphone-based multispectral imaging and analysis moreover offered better discrimination between seborrheic dermatitis and psoriasis than the RGB imaging and analysis. These results suggested that the multispectral imaging system based on a smartphone has the potential for self-diagnosis of seborrheic dermatitis with high portability and specificity.