Derivative works

This Defensive Disclosure document provides a series of technical descriptions for derivative inventions and improvements based on the core concepts disclosed in U.S. Patent 11,738,124. The purpose of this document is to place these concepts into the public domain, thereby establishing them as prior art for patentability purposes.

Disclosure 1: Material & Component Substitution

Derivative 1.1: Piezoresistive Microfluidic Anticoagulant Dosing System

• Enabling Description: This variation replaces the peristaltic anticoagulant pump (e.g., 234 in patent FIG. 3) with a solid-state microfluidic pump based on piezoresistive displacement. The pump consists of a silicon-on-glass substrate with a micro-etched channel and a chamber actuated by a PZT (lead zirconate titanate) ceramic element. A closed-loop control system uses a micro-flow sensor downstream of the pump to provide feedback to the controller. The controller adjusts the voltage and frequency applied to the PZT element to achieve a precise, non-pulsatile anticoagulant flow rate in the range of 0.1 to 15 mL/min with a resolution of 10 microliters. This substitution eliminates the mechanical wear and potential for spallation associated with peristaltic tubing and allows for a more accurate, real-time calculation of the infused anticoagulant volume, thereby improving the accuracy of the final pure plasma calculation.

• Diagram:

  graph TD
      subgraph Controller [Controller Unit]
          A[CPU] -- PZT Drive Signal --> B(PZT Actuator Driver);
          C(Micro-flow Sensor Interface) -- Flow Data --> A;
      end
      subgraph Microfluidic Pump
          D[Anticoagulant Reservoir] --> E{PZT-actuated Pump Chamber};
          E --> F[Micro-flow Sensor];
      end
      B --> E;
      F --> G((To Blood Draw Line));
      F -- Analog Signal --> C;
  

Derivative 1.2: Raman Spectroscopy for In-Line Hematocrit and Anticoagulant Sensing

• Enabling Description: This derivative replaces the separate hematocrit determination step and the reliance on pump rotation counting with a single, in-line Raman spectroscopy sensor. The sensor is positioned on the whole blood draw line, prior to the separation device. A 785 nm laser is focused through a sapphire window into the fluid path. The backscattered Raman-shifted photons are collected and analyzed by a spectrometer. The controller uses a chemometric model, pre-trained on samples with known hematocrit and anticoagulant (e.g., sodium citrate) concentrations, to simultaneously resolve the Raman peaks corresponding to hemoglobin (for hematocrit) and citrate ions. This provides a direct, real-time measurement of both donor hematocrit and the actual anticoagulant concentration in the anticoagulated whole blood, accounting for any pump inaccuracies or priming residues. This data feeds directly into the pure plasma calculation algorithm.

• Diagram:

  sequenceDiagram
      participant Donor
      participant RamanSensor as Raman Sensor
      participant Controller
      participant BloodSeparator as Separation Device
  
      Donor->>+RamanSensor: Whole Blood Flows
      RamanSensor->>RamanSensor: Excite with 785nm Laser
      RamanSensor->>Controller: Transmit Raman Spectrum
      Controller->>Controller: Apply Chemometric Model (Hct, [AC])
      Controller->>RamanSensor: Acknowledge
      RamanSensor-->>-BloodSeparator: Anticoagulated Blood Continues
      Controller->>BloodSeparator: Adjust Separation Parameters
  

Derivative 1.3: Tangential Flow Filtration with Graphene-Oxide Composite Membrane

• Enabling Description: The centrifugal blood component separation device is replaced with a tangential flow filtration (TFF) module. The core of this module is a hollow-fiber membrane cartridge composed of a polysulfone substrate with a functionalized surface layer of graphene-oxide (GO). The GO layer is engineered with a nominal pore size of 0.2 microns and a highly negative surface charge to repel cellular components (erythrocytes, leukocytes) while allowing plasma proteins to pass through. The system controller actively manages transmembrane pressure (TMP) and cross-flow velocity to minimize hemolysis and membrane fouling. The "pure plasma" calculation from the patent's core claim is used here to control the TFF process, terminating the collection once the target volume of cell-free, pure plasma (calculated by subtracting the measured anticoagulant volume from the total permeate volume) is achieved.

• Diagram:

  graph TD
      WB[Anticoagulated Whole Blood] --> P1(Pump);
      P1 --> TFF{TFF GO-Membrane Module};
      TFF -- Retentate (RBCs, etc.) --> R((To Donor/Waste));
      TFF -- Permeate (Plasma + AC) --> S[Permeate Weight Sensor];
      S --> C{Collection Bag};
      S -- Weight Data --> CTRL(Controller);
      CTRL -- Pump Control --> P1;
      CTRL --> R;
      subgraph Real-time Calculation
          CTRL -- "Pure Plasma Vol = Total Vol - AC Vol" --> Stop(Stop Collection?);
      end
  

Disclosure 2: Operational Parameter Expansion

Derivative 2.1: Nanoscale Apheresis for Exosome Isolation from Liquid Biopsy

• Enabling Description: The method is scaled down to a microfluidic lab-on-a-chip for isolating exosomes from microliter volumes of plasma for diagnostic purposes. A 100 µL plasma sample (already separated from whole blood) is introduced into the chip and mixed with an exosome precipitation agent (e.g., polyethylene glycol), which acts as the "anticoagulant" analog. The mixture flows into a separation chamber where acoustophoresis, using MHz-frequency standing surface acoustic waves, separates the precipitated exosomes from the bulk plasma. An in-line optical density sensor measures the concentration of the precipitation agent. The controller calculates the volume of "pure" exosome-free supernatant and routes the exosome concentrate to a collection well once a target supernatant removal volume is reached, ensuring a standardized exosome concentration for downstream analysis.

• Diagram:

  stateDiagram-v2
      [*] --> Priming
      Priming --> Sample_Injection: Chip Primed
      Sample_Injection --> Mixing: 100uL Plasma Injected
      Mixing: Mix with PEG
      Mixing --> Separation: Mixture Flows to Chamber
      Separation: Apply Acoustic Field
      Separation --> Sensing: Supernatant Flows Past Sensor
      Sensing --> Calculation: Measure PEG Concentration
      Calculation --> Decision
      Decision: Pure Supernatant Target Reached?
      Decision -- No --> Separation: Recirculate/Continue
      Decision -- Yes --> Collection
      Collection --> [*]
  

Derivative 2.2: Industrial Scale Bioreactor Harvest with Target Protein Calculation

• Enabling Description: The disclosed method is applied to a 2000-liter mammalian cell culture bioreactor for harvesting monoclonal antibodies (mAbs). A harvesting agent (e.g., a clarifying flocculant) is added to the bioreactor. The cell culture fluid is then passed through a depth filtration system. A downstream sensor (e.g., a turbidity or capacitance sensor) measures the concentration of the residual flocculant. The system controller calculates the volume of "pure" mAb-containing supernatant by subtracting the calculated flocculant volume from the total harvested volume. The harvest is automatically terminated when a pre-set target volume of pure supernatant is collected, ensuring consistent batch-to-batch dilution and optimizing downstream purification steps.

• Diagram:

  graph LR
      A[Bioreactor (2000L)] -- Add Flocculant --> B(Pump);
      B --> C{Depth Filtration Skid};
      C -- Cell Debris --> Waste;
      C -- Supernatant --> D[Flocculant Sensor];
      D --> E[Collection Tank];
      D -- Concentration Data --> F(Process Controller);
      F -- "Calculate Pure Supernatant Volume" --> F;
      F -- "IF Target Reached THEN Stop" --> B;
  

Disclosure 3: Cross-Domain Application

Derivative 3.1 (Aerospace): On-Orbit Water Purification Control

• Enabling Description: This system is adapted for a closed-loop water recycling system aboard a long-duration spacecraft. Wastewater is treated with a known concentration of an iodinated resin disinfectant, which acts as the "anticoagulant" analog. The water is then passed through a filtration and purification unit. A downstream ion-selective electrode (ISE) sensor measures the residual iodine concentration. The control system calculates the volume of "pure" potable water by subtracting the volume attributed to the disinfectant from the total processed volume. The system terminates the purification cycle for a given batch once a target volume of pure water is confirmed, ensuring compliance with purity standards for crew consumption.

• Diagram:

  flowchart TD
      A[Wastewater Tank] --> B{Dosing Unit};
      C[Iodine Resin] --> B;
      B --> D(Pump);
      D --> E[Purification Unit];
      E -- Brine/Waste --> F;
      E -- Processed Water --> G{Iodine ISE Sensor};
      G --> H[Potable Water Tank];
      G -- Iodine Conc. --> I(System Controller);
      I -- "Pure H2O Vol = Total Vol - Iodine Vol" --> J{Stop Cycle?};
      J -- Yes --> D;
  

Derivative 3.2 (AgTech): Automated Saffron Extraction Yield Optimization

• Enabling Description: The system is applied to the solvent-based extraction of crocin and safranal from saffron. Dried saffron stigmas are mixed with a specific volume of an ethanol-water solvent (the "anticoagulant" analog) in an extraction vessel. The mixture is then separated via filtration. A spectrophotometer in the filtrate line continuously measures the absorbance at specific wavelengths (e.g., 440 nm for crocin) and also the baseline shift caused by the solvent itself. The controller uses this data to calculate the real-time concentration and thus the total extracted mass of "pure" crocin, distinct from the solvent mass. The extraction process is dynamically controlled and stopped when the rate of increase of pure crocin yield falls below a threshold, optimizing vessel time and energy consumption.

• Diagram:

  sequenceDiagram
      participant Vessel as Extraction Vessel
      participant Spectro as Spectrophotometer
      participant Controller
      Vessel->>Vessel: Add Saffron & Solvent
      loop Extraction Process
          Vessel->>Spectro: Circulate Extract
          Spectro->>Controller: Transmit Absorbance Spectrum
          Controller->>Controller: Calculate Pure Crocin Yield
          Controller->>Controller: Analyze Yield Rate (dYield/dt)
          alt Rate > Threshold
              Controller-->>Vessel: Continue Extraction
          else Rate <= Threshold
              Controller-->>Vessel: Stop Extraction
              break
          end
      end
  

Disclosure 4: Integration with Emerging Technologies

Derivative 4.1 (AI Integration): Predictive AI for Donor-Specific Procedure Optimization

• Enabling Description: The system controller is augmented with a pre-trained Recurrent Neural Network (RNN) model. During the procedure, the model receives real-time inputs from IoT-enabled sensors: the Raman sensor (hematocrit), anticoagulant flow meter, blood pump pressure sensors, and a non-invasive cuff measuring the donor's heart rate and blood pressure. The RNN model predicts in real-time the donor's vascular response and potential for citrate reaction. It dynamically adjusts the blood draw rate and the anticoagulant-to-blood ratio to maximize the collection efficiency of pure plasma while keeping the donor's physiological parameters within a personalized safety envelope. The target pure plasma volume may be slightly adjusted downwards by the AI if it predicts an impending adverse reaction.

• Diagram:

  graph TD
      subgraph Inputs
          A[Donor Vitals (HR, BP)];
          B[Hematocrit Sensor];
          C[Pressure Sensor];
          D[AC Flow Rate];
      end
      subgraph Controller
          E(RNN Model);
          F[Control Logic];
      end
      subgraph Outputs
          G[Blood Pump Speed];
          H[AC Pump Speed];
          I[Adjusted Target Volume];
      end
      A & B & C & D --> E;
      E -- Predictions --> F;
      F --> G & H & I;
  

Derivative 4.3 (Blockchain Verification): Immutable Ledger for Plasma Provenance

• Enabling Description: Upon completion of a donation, the apheresis machine's controller generates a data block containing the final calculated pure plasma volume, the donor's anonymized cryptographic ID, the timestamp, the machine's unique identifier, the disposable kit's unique ID (read from an RFID tag), and a hash of the full procedure log. This data block is signed with the machine's private key and submitted as a transaction to a permissioned Hyperledger Fabric blockchain shared between collection centers, testing labs, and fractionation facilities. This creates an auditable, immutable record for each unit of plasma, ensuring full traceability and verifying that the collected volume corresponds to the "pure plasma" standard, not the total fluid volume.

• Diagram:

  erDiagram
      DONOR {
          string anonymousDonorID PK
      }
      PLASMA_UNIT {
          string unitID PK
          string anonymousDonorID FK
          string machineID FK
          float purePlasmaVolume
          datetime timestamp
          string transactionHash
      }
      MACHINE {
          string machineID PK
          string location
      }
      PLASMA_UNIT ||--o{ DONOR : collectedFrom
      PLASMA_UNIT ||--o{ MACHINE : collectedOn
  

Disclosure 5: The "Inverse" or Failure Mode

Derivative 5.1: Graceful Degradation to Volumetric Limiting on Sensor Failure

• Enabling Description: The system is designed with a safe-fail mode for the primary hematocrit sensor (e.g., Raman sensor or optical sensor). The controller continuously performs a self-check on the sensor's signal, looking for values outside an expected physiological range or a lack of signal variance. If a sensor failure is detected, the controller triggers three actions: 1) It immediately flags an alert to the operator's user interface. 2) It transitions its control algorithm from "pure plasma target" to a conservative "total volume target". This new target is calculated as the lesser of either the FDA-allowed total volume (e.g., 880 mL) or 110% of the originally intended pure plasma target (e.g., 110% of 800 mL = 880 mL). 3) It logs the failure event and the exact time of the control mode switch. This ensures the procedure can be completed safely without risking over-collection, prioritizing donor safety over yield optimization when critical data is unavailable.

• Diagram:

  stateDiagram-v2
      state "Normal Operation (Pure Plasma Target)" as Normal {
          [*] --> Running
          Running: HCT_Sensor_OK
          Running --> Sensor_Fail_Detected : event SensorFailure
      }
      state "Degraded Mode (Total Volume Target)" as Degraded {
          Entry: Alert Operator
          Entry: Set Conservative Target
          [*] --> Running
          Running: Collect based on total volume
          Running --> Procedure_Complete
      }
      Normal --> Degraded : HCT_Sensor_Fail
      Degraded --> [*]
  

Combination Prior Art Scenarios

1. AI On-Device with TensorFlow Lite: The predictive AI model for donor-specific optimization (Derivative 4.1) is compiled using the TensorFlow Lite open-source machine learning framework. The resulting model is deployed directly onto the apheresis machine's embedded ARM-based controller. This enables real-time, low-latency inference for pump and ratio adjustments without requiring a constant cloud connection, making the advanced optimization feature robust and self-contained.

2. IoT Data Streaming with MQTT: The IoT sensors monitoring the apheresis procedure (Derivative 4.2) are configured to publish their data (pressure, temperature, flow rates, status) to a central broker using the MQTT (Message Queuing Telemetry Transport) protocol, an ISO standard (ISO/IEC 20922). This lightweight pub/sub protocol ensures reliable data delivery over potentially unreliable networks (like Wi-Fi in a busy clinic), allowing a remote fleet management dashboard to monitor thousands of devices efficiently.

3. Blockchain Provenance with Hyperledger Fabric: The system for creating an immutable ledger of plasma provenance (Derivative 4.3) is implemented using the open-source Hyperledger Fabric framework. A "chaincode" (smart contract) is written to define the transaction structure (donor ID, pure plasma volume, machine ID, etc.) and the business logic for validating and committing new entries to the blockchain. This leverages a widely adopted, enterprise-grade open-source standard for building the distributed ledger, ensuring interoperability and security.