By John Vandermosten, CFA

NASDAQ:MDAI

READ THE FULL MDAI RESEARCH REPORT

Spectral AI, Inc. (NASDAQ:MDAI) detailed 2024 financial and operational results in a March 27^th press release. Revenues of $29.6 million exceeded guidance of $28 million and our estimates of $29.3 million. Net loss was ($15.3) million or ($0.85) per share. Over the last months, Spectral has been operating at a rapid pace with topline data reported from its burn validation study in a March 17^th press release confirming DeepView’s superior detection sensitivity relative to burn physicians’ assessments. Data regarding DeepView’s sensitivity, specificity and Dice Score were provided. We expect an increased allocation of effort towards the submission of the De Novo application to the FDA, which is slated for mid-2025. The company has also raised additional capital, installed three devices in Australia and prepared for the intended spin-off of Spectral IP.

2024 Financial and Operational Results

Spectral reported 2024 results in a press release on March 27^th followed by a conference call to discuss results with investors. A Form 10-K was subsequently filed with the SEC. For the year ending December 31^st, 2024, research and development revenues of $29.6 million were recognized. Net loss for 2024 totaled ($15.3) million or ($0.85) per share. For 2024 versus the same prior year period in 2023:

• Revenues were $29.6 million up 64% from $18.1 million due to an increase in activity related to BARDA and other US government grant revenues. BARDA revenues totaled $19.7 million while other government contracts composed primarily of MTEC funds were $1.8 million;
• Cost of revenue, which can be thought of as research and development expense, totaled $16.3 million, rising 60% from $10.2 million due to higher activity levels related to the BARDA and MTEC contracts. Gross margin improved to 44.9% from 43.6% due to a higher reimbursement rate in the new BARDA contract;
• General & Administrative expenses were $19.9 million, down 5% from $20.9 million reflecting lower legal and accounting expenses as well as lower overall spending;
• Net interest income was $14,000 compared to $172,000;
• Other expense was ($8.5) million vs. ($8.0) million with borrowing related costs and change in fair value of warrants replacing transaction costs;
• Income tax was ($271,000) vs. ($11,000);
• Net loss was ($15.3) million vs. ($20.9) million or ($0.85) and ($1.48) per share, respectively.

As of December 31^st, 2024, cash totaled $5.2 million. This amount compares to the $4.8 million cash balance held at the end of 2023. Several financing transactions took place in 2024 including contributions from issuance of common stock and notes offset by repayment of other notes. The company filed a shelf registration on Form S-3 to raise additional capital in 4Q:24. Days prior to the 2024 report, Spectral announced that it had raised about $11.2 million gross in a debt and equity financing along with access to additional debt of $6.5 million upon reaching FDA clearance of DeepView.

March 2025 Financing

Spectral AI accessed gross proceeds of approximately $11.2 million in a March 2025 financing that was funded with both debt and equity. Avenue Venture Opportunities Fund provided an initial draw down of $8.5 million in a $15 million debt facility. Equity financing was $2.7 million contributed from existing and institutional investors. Total cash following the debt and equity financing is in excess of $14 million. Spectral has access to another $6.5 million in debt following FDA clearance of DeepView. Capital will be used for DeepView commercialization. The term of the financing is three years with an interest-only payment period of fifteen months. Avenue received warrants with an exercise price of $1.80, equal to 8.5% of the funding commitment. The equity raise sold 2,076,923 shares of the company’s common stock at $1.30 per share to raise $2.7 million. Each share has a warrant attached with an exercise price of $1.80. There is a second financing tranche that is contingent on regulatory clearance. This would provide $6.5 million of additional debt financing and require an additional $7 million equity raise by the company.

Registration Statement Filed for Spectral IP IPO

Spectral AI’s wholly-owned subsidiary, Spectral IP, Inc., filed a registration statement with the SEC related to the spinoff and an initial public offering of its shares of common stock. The new company will be renamed SIM IP, Inc. The number of shares and pricing have not yet been determined. Dominari Securities is acting as the representative of the underwriters and Revere Securities is acting as the co-underwriter for the proposed offering. Further details of the offering are available in the March 20^thpress release and in Form S-1.

The company will focus on IP-based financing and monetization strategies including licensing, litigation investment, structured financing, royalty acquisition and opportunistic transactions. It will offer an industry agnostic investment mandate but will focus on healthcare, semiconductors, AI and internet of things (IOT) industries where intellectual property is a critical component of product success. IP related to Spectral AI and its imaging systems will remain with Spectral AI. Ownership in the entity, assuming the company is publicly offered, will be distributed to Spectral AI shareholders. The transaction is expected to be complete in the next three months.

Burn Validation Study Results

On February 7^th, Spectral announced that it had completed the data analysis of the burn validation study presaging the release of topline data on March 17^th. Results comparing the sensitivity, Dice score and specificity were provided along with anticipated timing for clearance and commercialization.

Burn Trial Background

On January 11^th, 2024, Spectral announced that it began enrollment of its pivotal study to validate DeepView for burn injuries. The study enrolled and analyzed data from 164 patients, consisting of 115 adults and 49 children in burn centers and emergency departments throughout the US. The study was filed on clinicaltrials.gov under the designator NCT06131203 and is titled Burn Pivotal Study. Its purpose is to validate the burn healing assessment algorithm for the DeepView device. 16 sites enrolled patients.

Analysis of study data generated results related to the sensitivity, specificity and Dice score of the DeepView system at both a pixel level and full image level compared with the clinical judgment of burn physicians. Findings of the various endpoints were:

• Sensitivity - The DeepView System demonstrated a statistically significant improvement in identifying non-healing tissue compared to burn physicians, as judged on sensitivity. At the image-wise level, the DeepView System scored 86.6% while clinical judgment annotation (CJA) of burn physicians scored 40.8%. At the pixel-wise level, the DeepView System scored 81.9% and CJA of burn physicians scored 38.8%;
• Dice Score^[1] - The DeepView System achieved statistically significant higher Dice Scores when compared to those derived from burn physicians’ CJA, representing the improved pixel-wise evaluation between predicated and true segmented wound areas with the DeepView System performing at 68.5% and burn physician’s CJA at 39.2%;
• Specificity - The DeepView System produced a specificity of 61.2% (versus an anticipated result of 36.0%) and CJA of 79.1%.

The anticipated regulatory pathway will include a De Novo FDA submission for DeepView in June 2025, which, if successfully approved, will allow for commercialization of the device in 2026. Spectral will request consideration of DeepView as a Class II device.^[2] After receiving the approval for use in burn centers, Spectral will pursue emergency department approval via the 510(k) pathway.^[3]

Beyond the binary output provided by the DeepView Burn diagnostic, the device is also able to rapidly and accurately determine burn size. Details of a proof-of-concept module was detailed in an October 3, 2024 press release. In seconds, DeepView Burn can provide the total body surface area (TBSA) of a burn and calculate depth, area and volume assisting providers to determine the best course forward. This technology is embedded into the DeepView system and can aid clinicians in their efforts to improve patient treatment decisions. We believe that this functionality will be part of the FDA submission and formalized in the instructions for use.

SUBSCRIBE TO ZACKS SMALL CAP RESEARCH to receive our articles and reports emailed directly to you each morning. Please visit our website for additional information on Zacks SCR. 

DISCLOSURE: Zacks SCR has received compensation from the issuer directly, from an investment manager, or from an investor relations consulting firm, engaged by the issuer, for providing research coverage for a period of no less than one year. Research articles, as seen here, are part of the service Zacks SCR provides and Zacks SCR receives quarterly payments totaling a maximum fee of up to $40,000 annually for these services provided to or regarding the issuer. Full Disclaimer HERE.

________________________

^{[1] The Dice Score, also known as the Dice Similarity Coefficient (DSC), is a metric used to measure the similarity between two sets, commonly used in image segmentation tasks to compare the predicted segmentation with the ground truth. It is commonly used in medical image segmentation, computer vision and natural language processing. The DSC contrasts with simple accuracy metrics like sensitivity and specificity in that it considers both size and spatial alignment of segmented regions, reflecting perceptual quality more effectively.}

^{[2] Class II medical devices require greater regulatory controls than Class I devices to ensure safety and effectiveness but are typically less risky than Class III devices. The FDA classifies medical devices based on the level of control necessary to provide a reasonable assurance of the device's safety and effectiveness. Class II medical devices present moderate risk and can pose a moderate risk to the patient if used incorrectly or malfunction. They require special controls in addition to the general controls required for Class I devices which may include specific labeling requirements, mandatory performance standards and post-market surveillance.}

^{[3] The 510(k) pathway is designed for devices that are substantially equivalent to a legally marketed device known as a predicate device. The device must not pose new risks. 510(k) is primarily intended for Class II devices, but some low-risk Class I devices also follow this pathway. The approval process calls for the manufacturer to demonstrate that the device is similar in safety and effectiveness to an already-approved device. Clinical trials are usually not required unless substantial changes are made to the design or use.}